Đánh giá khả năng tăng cường dữ liệu trong dự báo khả năng chịu cắt của dầm bê tông cốt thép được tăng cường bởi bê tông UHPC
Email:
hoangviethai@utc.edu.vn
Từ khóa:
Tăng cường dữ liệu, Khả năng chịu cắt, UHPC, Mô hình học máy.
Tóm tắt
Việc dự báo khả năng chịu cắt của dầm bê tông cốt thép (BTCT) được tăng cường bằng lớp bê tông siêu tính năng (UHPC) đang là hướng nghiên cứu thu hút nhiều quan tâm hiện nay. Tuy nhiên, dữ liệu thực nghiệm còn rất hạn chế do chi phí cao. Nghiên cứu này đề xuất phương án tăng cường dữ liệu dựa trên phân phối Gaussian nhằm nâng cao hiệu suất dự báo sức kháng cắt của dầm BTCT được tăng cường bởi UHPC. Bốn mô hình học máy bao gồm: Random Forest (RF), KNN, LightGBM và XGBoost đã được sử dụng để đánh giá. Quá trình kiểm tra được thực hiện hoàn toàn trên tập dữ liệu thực nghiệm gốc để đảm bảo đánh giá khách quan khả năng khái quát hóa. Kết quả cho thấy các mô hình chỉ huấn luyện trên dữ liệu thực nghiệm bị hạn chế và có sai số khá lớn. Sau khi áp dụng tăng cường dữ liệu, ngoại trừ mô hình KNN, hiệu suất của các mô hình khác được cải thiện rõ rệt. Mô hình XGBoost đạt hiệu quả cao nhất với R2 kiểm tra lên tới 0,949, MAE = 36,558 kN và RMSE = 54,737 kN. Kết quả này giúp giải quyết hiệu quả tình trạng khan hiếm dữ liệu. Cách tiếp cận này cung cấp một phương án tin cậy cho việc đánh giá và thiết kế tăng cường các dầm BTCT bằng bê tông UHPC.Tài liệu tham khảo
[1]. V.H. Hoang, T.A Do, A. T. Tran, X. H Nguyen, Flexural capacity of reinforced concrete slabs retrofitted with ultra-high-performance concrete and fiber-reinforced polymer, Innovative Infrastructure Solutions, 9 (2024). https://doi.org/10.1007/s41062-024-01410-y.
[2]. L. Liu, S. Wan, Flexural bearing capacity of reinforced concrete beams reinforced with carbon fiber reinforced plastics strips and ultra-high performance concrete layers, International Journal of Building Pathology and Adaptation, (2022). https://doi.org/10.1108/IJBPA-04-2022-0056
[3]. Y. Zhang, X. Li, Y. Zhu, X. Shao, Experimental study on flexural behavior of damaged reinforced concrete (RC) beam strengthened by toughness-improved ultra-high-performance concrete (UHPC) layer, Composites Part B: Engineering, 186 (2020). https://doi.org/10.1016/j.compositesb.2020.107834
[4]. Eurocode 2, Design of concrete structures – Part 1: General rules and rules for buildings, Brussels, Belgium, 1992.
[5]. H.R.M. Mohammed, S. Ismail, Proposition of new computer artificial intelligence models for shear strength prediction of reinforced concrete beams, Engineering with Computers, 38 (2022) 3739–3757. https://doi.org/10.1007/s00366-021-01400-z
[6]. A. Kumar, H.C. Arora, N.R. Kapoor et al., Machine learning intelligence to assess the shear capacity of corroded reinforced concrete beams, Scientific Reports, 13 (2023) 2857. https://doi.org/10.1038/s41598-023-30037-9
[7]. W.Z. Taffese, Y. Zhu, Explainable machine learning for predicting flexural capacity of reinforced UHPC beams, Engineering Structures, 343 (2025) 121188. https://doi.org/10.1016/j.engstruct.2025.121188
[8]. V.H. Hoang, M.Q. Tran, V.T. Ngo, Machine learning-based prediction of the axial load capacity of UHPC strengthened reinforced concrete columns: A comparative analysis, PLOS ONE, 21 (2026) e0338120. https://doi.org/10.1371/journal.pone.0338120
[9]. W.Z. Taffese, N. Khodadadi, Y. Zhu, S. Mirjalili, A. Nanni, A generative adversarial network enhanced ensemble learning-based prediction model for moment improvement effect of UHPC strengthened damaged RC beams, Case Studies in Construction Materials, 23 (2025) e05323. https://doi.org/10.1016/j.cscm.2025.e05323
[10]. C. Chokwitthaya, Y. Zhu, S. Mukhopadhyay, A. Jafari, Applying the Gaussian Mixture Model to generate large synthetic data from a small dataset, Conference Proceeding in Construction Research Congress 2020: Computer Applications (2020).
[11]. T.M. Tran, H.L. Le, Enhanced prediction of the flexural capacity of prestressed reinforced concrete beams using an improved PSO-ANN model, Engineering, Technology & Applied Science Research, 16 (2026) 31947–31953. https://doi.org/10.48084/etasr.15746
[12]. M.A. Sakr, A.A. Sleemah, T.M. Khalifa, W.N. Mansour, Shear strengthening of reinforced concrete beams using prefabricated ultra-high performance fiber reinforced concrete plates: Experimental and numerical investigation, Structural Concrete, 20 (2019) 1137–1153. https://doi.org/10.1002/suco.201800137
[13]. A.A. Bahraq, et al., Experimental and numerical investigation of shear behavior of RC beams strengthened by ultra-high-performance concrete, International Journal of Concrete Structures and Materials, 13 (2019) 6. https://doi.org/10.1186/s40069-018-0330-z
[14]. H. Ji, C. Liu, Ultimate shear resistance of ultra-high performance fiber reinforced concrete-normal strength concrete beam, Engineering Structures, 203 (2020) 109825. https://doi.org/10.1016/j.engstruct.2019.109825
[15]. A. Sine, M. Pimentel, S. Nunes, A. Dimande, Shear behaviour of RC-UHPFRC composite beams without transverse reinforcement, Engineering Structures, 257 (2022) 114053. https://doi.org/10.1016/j.engstruct.2022.114053
[16]. L. Ke, et al., Shear performance evaluation of damaged RC beams strengthened with cast-in-place U-shaped UHPFRC shell, Structures, 58 (2023) 105530. https://doi.org/10.1016/j.istruc.2023.105530
[17]. S.-G. Hong, W.-Y. Lim, Strengthening of shear-dominant reinforced concrete beams with ultra-high-performance concrete jacketing, Construction and Building Materials, 365 (2023) 130043. https://doi.org/10.1016/j.conbuildmat.2022.130043
[18]. X. Liu, G.E. Thermou, Shear performance of RC beams strengthened with high-performance fibre-reinforced concrete (HPFRC) under static and fatigue loading, Materials, 17 (2024) 5227. https://doi.org/10.3390/ma17215227.
[19]. A. Abd Elghany, M. Elsayed, A.A. Elsayed, A. Shaheen, Enhancement of the shear capacity of RC deep beams with ultra-high performance fiber-reinforced concrete, Engineering, Technology & Applied Science Research, 15 (2025) 20418–20424. https://doi.org/10.48084/etasr.9792
[20]. L. Breiman, Random forests, Machine Learning, 45 (1) (2001) 5–32.
[21]. Z.J. Zhang, Introduction to machine learning: k-nearest neighbors, Annals of Translational Medicine, 4 (11) (2016) 218. http://dx.doi.org/10.21037/atm.2016.03.37
[22]. J.B. Brownlee, XGBoost with Python: Gradient boosted trees with XGBoost and scikit-learn, Machine Learning Mastery, (2016) 115. https://machinelearningmastery.com/xgboost-with-python/
[23]. G. Ke, Q. Meng, T. Finley, T. Wang, W. Chen, W. Ma, et al., LightGBM: a highly efficient gradient boosting decision tree, in: Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS), 2017.
[2]. L. Liu, S. Wan, Flexural bearing capacity of reinforced concrete beams reinforced with carbon fiber reinforced plastics strips and ultra-high performance concrete layers, International Journal of Building Pathology and Adaptation, (2022). https://doi.org/10.1108/IJBPA-04-2022-0056
[3]. Y. Zhang, X. Li, Y. Zhu, X. Shao, Experimental study on flexural behavior of damaged reinforced concrete (RC) beam strengthened by toughness-improved ultra-high-performance concrete (UHPC) layer, Composites Part B: Engineering, 186 (2020). https://doi.org/10.1016/j.compositesb.2020.107834
[4]. Eurocode 2, Design of concrete structures – Part 1: General rules and rules for buildings, Brussels, Belgium, 1992.
[5]. H.R.M. Mohammed, S. Ismail, Proposition of new computer artificial intelligence models for shear strength prediction of reinforced concrete beams, Engineering with Computers, 38 (2022) 3739–3757. https://doi.org/10.1007/s00366-021-01400-z
[6]. A. Kumar, H.C. Arora, N.R. Kapoor et al., Machine learning intelligence to assess the shear capacity of corroded reinforced concrete beams, Scientific Reports, 13 (2023) 2857. https://doi.org/10.1038/s41598-023-30037-9
[7]. W.Z. Taffese, Y. Zhu, Explainable machine learning for predicting flexural capacity of reinforced UHPC beams, Engineering Structures, 343 (2025) 121188. https://doi.org/10.1016/j.engstruct.2025.121188
[8]. V.H. Hoang, M.Q. Tran, V.T. Ngo, Machine learning-based prediction of the axial load capacity of UHPC strengthened reinforced concrete columns: A comparative analysis, PLOS ONE, 21 (2026) e0338120. https://doi.org/10.1371/journal.pone.0338120
[9]. W.Z. Taffese, N. Khodadadi, Y. Zhu, S. Mirjalili, A. Nanni, A generative adversarial network enhanced ensemble learning-based prediction model for moment improvement effect of UHPC strengthened damaged RC beams, Case Studies in Construction Materials, 23 (2025) e05323. https://doi.org/10.1016/j.cscm.2025.e05323
[10]. C. Chokwitthaya, Y. Zhu, S. Mukhopadhyay, A. Jafari, Applying the Gaussian Mixture Model to generate large synthetic data from a small dataset, Conference Proceeding in Construction Research Congress 2020: Computer Applications (2020).
[11]. T.M. Tran, H.L. Le, Enhanced prediction of the flexural capacity of prestressed reinforced concrete beams using an improved PSO-ANN model, Engineering, Technology & Applied Science Research, 16 (2026) 31947–31953. https://doi.org/10.48084/etasr.15746
[12]. M.A. Sakr, A.A. Sleemah, T.M. Khalifa, W.N. Mansour, Shear strengthening of reinforced concrete beams using prefabricated ultra-high performance fiber reinforced concrete plates: Experimental and numerical investigation, Structural Concrete, 20 (2019) 1137–1153. https://doi.org/10.1002/suco.201800137
[13]. A.A. Bahraq, et al., Experimental and numerical investigation of shear behavior of RC beams strengthened by ultra-high-performance concrete, International Journal of Concrete Structures and Materials, 13 (2019) 6. https://doi.org/10.1186/s40069-018-0330-z
[14]. H. Ji, C. Liu, Ultimate shear resistance of ultra-high performance fiber reinforced concrete-normal strength concrete beam, Engineering Structures, 203 (2020) 109825. https://doi.org/10.1016/j.engstruct.2019.109825
[15]. A. Sine, M. Pimentel, S. Nunes, A. Dimande, Shear behaviour of RC-UHPFRC composite beams without transverse reinforcement, Engineering Structures, 257 (2022) 114053. https://doi.org/10.1016/j.engstruct.2022.114053
[16]. L. Ke, et al., Shear performance evaluation of damaged RC beams strengthened with cast-in-place U-shaped UHPFRC shell, Structures, 58 (2023) 105530. https://doi.org/10.1016/j.istruc.2023.105530
[17]. S.-G. Hong, W.-Y. Lim, Strengthening of shear-dominant reinforced concrete beams with ultra-high-performance concrete jacketing, Construction and Building Materials, 365 (2023) 130043. https://doi.org/10.1016/j.conbuildmat.2022.130043
[18]. X. Liu, G.E. Thermou, Shear performance of RC beams strengthened with high-performance fibre-reinforced concrete (HPFRC) under static and fatigue loading, Materials, 17 (2024) 5227. https://doi.org/10.3390/ma17215227.
[19]. A. Abd Elghany, M. Elsayed, A.A. Elsayed, A. Shaheen, Enhancement of the shear capacity of RC deep beams with ultra-high performance fiber-reinforced concrete, Engineering, Technology & Applied Science Research, 15 (2025) 20418–20424. https://doi.org/10.48084/etasr.9792
[20]. L. Breiman, Random forests, Machine Learning, 45 (1) (2001) 5–32.
[21]. Z.J. Zhang, Introduction to machine learning: k-nearest neighbors, Annals of Translational Medicine, 4 (11) (2016) 218. http://dx.doi.org/10.21037/atm.2016.03.37
[22]. J.B. Brownlee, XGBoost with Python: Gradient boosted trees with XGBoost and scikit-learn, Machine Learning Mastery, (2016) 115. https://machinelearningmastery.com/xgboost-with-python/
[23]. G. Ke, Q. Meng, T. Finley, T. Wang, W. Chen, W. Ma, et al., LightGBM: a highly efficient gradient boosting decision tree, in: Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS), 2017.
Tải xuống
Chưa có dữ liệu thống kê
Nhận bài
26/01/2026
Nhận bài sửa
12/03/2026
Chấp nhận đăng
17/03/2026
Xuất bản
15/04/2026
Chuyên mục
Công trình khoa học
Kiểu trích dẫn
Lê Hà, L., Ngô Đức, C., Lê Đắc, H., & Hoàng Việt, H. (1776186000). Đánh giá khả năng tăng cường dữ liệu trong dự báo khả năng chịu cắt của dầm bê tông cốt thép được tăng cường bởi bê tông UHPC. Tạp Chí Khoa Học Giao Thông Vận Tải, 77(3), 297-312. https://doi.org/10.47869/tcsj.77.3.6
